PhD defense by Marta Agnieszka Mrozowska

Title: Boundary Layer Turbulence: Exploring Strategies to Improve Mixing Parameterizations in Ocean Models

Abstract: 
The oceanic boundary layer is the barrier connecting the ocean interior to other climate components. Simultaneously, it is the region in the sea with the highest turbulence density. In global ocean simulations, turbulent fluxes are too small to resolve and must therefore be parameterized.

Vertical mixing schemes are thus integral for accurate representation of upper ocean dynamics in climate models. However, persistent surface biases continue to limit the accuracy of numerical climate simulations. In this thesis, two strategies to improve oceanic vertical mixing schemes are explored. The first study investigates whether a range of parameterizations can reproduce the observed near-inertial wave-induced mixing at two sites in the Tropical Atlantic. Shipboard turbulence observations are compared to two forced, eddy-rich ocean model simulations. The observed mixing is not reproduced in any of the models, but near-inertial wave amplitude is found to be sensitive to parameterization choice. In the second study, Bayesian optimization as a method for automated tuning of ocean models is proposed. The Turbulent Kinetic Energy (TKE) scheme is tuned to minimize mixed layer depth (MLD) biases in the ocean model Veros. The default TKE parameter values fall within the parameter space region for which MLD bias is minimized.

Chairman: Prof. Klaus Mosegaard

Committee: Prof. Dr. Hans Burchard and Prof. Fabien Rouquet

Supervisor: Prof. Markus Jochum and Assoc. Prof. James Avery